If all four of the battery breakthrough articles on that page actually worked in a product, battery performance would be far higher than it is now. It seems to be possible to trade off charging rate, Wh/L, Wh/Kg, number of cycles, and safety. Any article that doesn't give all the stats is deceptive.
It's progress. The trouble is reading about it through the hype department at the university's PR operation.
Gunpowder is 4-10 times less energy dense than gasoline. The difference is that gunpowder includes fuel and oxygen-producing substances, much like most of Li-ion batteries.
A good hybrid can do very well. Presumably by keeping the engine in exactly its sweet spot and designing aggressively for that. BYD for example claims 46% thermal efficiency. [0]
Batteries can make use of the atmosphere as well (eg aluminum air batteries/ or venting hydrogen in lead acid batteries) although I don’t know of any rechargeable chemistries off hand that use environmental oxygen. All that to say the trick is available for batteries even if the best current chemistries by mass density don’t make use of it.
A good recycling program sounds like a tall order. I'm seeing Silver nanoparticles (heavy metal) and multiple things that react violently with water.
I'm always skeptical of any idea that ends with a bespoke industrial-scale recycling process. People tend to massively underestimate the complexity of recycling, especially at scale.
In general, bespoke recycling processes can make sense, especially if you manage to design the items to recycle with the recycling process in mind. There are several types of goods where this is put into practice (paper, compounds like TetraPak packages, various polymer plastics). Not sure about all the differrent types of batteries, though.
That's only a valid concept in some embedded engineering case, where a certain capacity is required, and double that amount is provisioned to account for degradation.
Few consumers think this way. Something doesn't have double the capacity that it has; the capacity is the capacity, and the decline looks bad.
The whole idea of the embedded part is that you make the degredation invisible to the consumer for as long as possible. From the factory, only charge up to ~4.07 Volts or thereabouts. Every N cycles, add 0.01 V to the threshold. Your phone probably already does something like this.
But yeah, 20% degredation in 100 cycles is atrocious. No amount of firmware shenanigans will be able to paper over that, not in any regular consumer product at least.
I can still think of use cases, though. Reserve power sources that aren't meant to be cycled daily, where smallness is valuable. Those little car jumper packs, for example. If there was a UPS close to the size of a regular power strip, I'd buy a few.
> Your phone probably already does something like this.
It most certainly does not. Most devices track battery health % (last full capacity divided by design capacity) and the gauge just presents state of charge (current capacity/lastfull)
The better phone charge threshold systems measure usage and keep the phone in the 30-80% soc range as often as possible.
Voltage drops faster on old cells as they age so you need a coulomb counter. Only extremely shit designs guess soc based on voltage alone.
Engineering is compromise though. If you can make a hybrid that loses 5% at 100 but still retains 500wh/l you’re in good shape.
There was someone working on a membrane a while back that’s pretty good at diffusing the lithium transfer in a way that reduces dendrite formation substantially, for instance. That’ll drop your volumetric advantage and likely your max discharge and charge rate a bit but would fix a lot of other problems in the bargain.
I’m not saying that the solution, but there is a palette of tools you can mix and match and that may be one of them.
I saw a video on the CATL sodium batteries the other day and the deal is that they’ve found a way to reinforce the material in a way that brings up the slope of the back half of the discharge curve so it’s almost as good as lithium down to about 20% state of charge before falling off the cliff. Lithium is more like 10% but that’s something you can manage with charge circuitry and overprovisioning.
So yeah I’d like to know the answer to your question too.
Not if your application requires 2X the energy. Aircraft, drones, etc. There's always trade-offs in battery design. As an old saying goes: you can have high specific energy, low degradation, or low cost... pick two!
Charge cycle capacity drops are generally not linear. If we start with 2x capacity and drop to 1.6x after 100 cycles, then we might end up with 1.2x after 1000 cycles. Some smartphone manufacturers would love that as you start with extremely superior energy density and then have a built-in obsolescence.
LFP are common in EV’s and ‘solar generator’ style battery packs, but I’ve never seen them in phones or laptops (outside OLPC), reduced capacity makes them not great in these, better to go NMC.
No one knows, the paper just focused on 100 cycles, but it suggests that if its good at 100 it probably is not terrible at further cycles. I guess we'll have to wait for the next paper but the conclusion seems optimistic about future research:
It is important to note that additional improvements in practical cell parameters, such as further optimized electrolyte (E/C ratio), increased stack pressure, optimized separator selection, and higher areal capacity of cathodes, can potentially enhance both the energy density and cycling performance beyond laboratory-scale demonstrations.
Post-mortem analyses confirmed reduced Li accumulation, minimized swelling, and suppressed cathode degradation, validating the robust interfacial stability of the system. By concurrently addressing the reversibility of Li metal and the structural stability of Ni-rich layered cathodes, this synergistic design offers a scalable and manufacturable pathway toward high-energy, long-life anode-free LMBs.
A battery cell is a long thin ribbon that is rolled into a spiral shape. There's no way you can apply any mechanical agitation to all the layers. It's been tried, but nothing came out of it.
If all four of the battery breakthrough articles on that page actually worked in a product, battery performance would be far higher than it is now. It seems to be possible to trade off charging rate, Wh/L, Wh/Kg, number of cycles, and safety. Any article that doesn't give all the stats is deceptive.
It's progress. The trouble is reading about it through the hype department at the university's PR operation.
> hype department
Yes but:
It's signal that battery tech will maintain its cost-learning-curve for some time to come.
It'll be noteworthy, to me, once these announcements start to trail off.
For comparison gasoline has about 9000 Wh/L of raw chemical energy, of which maybe 30-40% gets converted to useful work.
https://en.wikipedia.org/wiki/Energy_density
Gunpowder is 4-10 times less energy dense than gasoline. The difference is that gunpowder includes fuel and oxygen-producing substances, much like most of Li-ion batteries.
This thing is in gunpowder energy density range.
30…40% is very ideal number, 15…25% is often the reality.
A good hybrid can do very well. Presumably by keeping the engine in exactly its sweet spot and designing aggressively for that. BYD for example claims 46% thermal efficiency. [0]
[0] https://ev.com/news/byd-hybrid-efficiency
Almost half way there
ICE engines outsource half of the reactants to the atmosphere, so this comparison isn't as useful as it appears at first.
Batteries can make use of the atmosphere as well (eg aluminum air batteries/ or venting hydrogen in lead acid batteries) although I don’t know of any rechargeable chemistries off hand that use environmental oxygen. All that to say the trick is available for batteries even if the best current chemistries by mass density don’t make use of it.
> the battery retained 81.9% of its initial capacity after 100 cycles
That's really terrible.
It's interesting, but 20% loss after 100 cycles is just not great. NMC gets that at near 1000 cycles. LFP gets that at near 5000 cycles.
Seemingly adequate for certain drone applications like in Ukraine. They may only need a couple charge cycles, and 4x the capacity is huge.
20% loss isn't too bad if you start out at double the capacity though.
My first thought was put the new cells in aircraft, then cheap cars finally grid storage
That actually could make sense especially with a good recycling program. Swap the packs every flight and recycle anything that falls below standards.
A good recycling program sounds like a tall order. I'm seeing Silver nanoparticles (heavy metal) and multiple things that react violently with water.
I'm always skeptical of any idea that ends with a bespoke industrial-scale recycling process. People tend to massively underestimate the complexity of recycling, especially at scale.
In general, bespoke recycling processes can make sense, especially if you manage to design the items to recycle with the recycling process in mind. There are several types of goods where this is put into practice (paper, compounds like TetraPak packages, various polymer plastics). Not sure about all the differrent types of batteries, though.
Surely there's a well trod progression, no? Something like military, space, drones, aircraft, IoT, consumer (phones, watches), vehicle, residential, grid?
That's only a valid concept in some embedded engineering case, where a certain capacity is required, and double that amount is provisioned to account for degradation.
Few consumers think this way. Something doesn't have double the capacity that it has; the capacity is the capacity, and the decline looks bad.
The whole idea of the embedded part is that you make the degredation invisible to the consumer for as long as possible. From the factory, only charge up to ~4.07 Volts or thereabouts. Every N cycles, add 0.01 V to the threshold. Your phone probably already does something like this.
But yeah, 20% degredation in 100 cycles is atrocious. No amount of firmware shenanigans will be able to paper over that, not in any regular consumer product at least.
I can still think of use cases, though. Reserve power sources that aren't meant to be cycled daily, where smallness is valuable. Those little car jumper packs, for example. If there was a UPS close to the size of a regular power strip, I'd buy a few.
> Your phone probably already does something like this.
It most certainly does not. Most devices track battery health % (last full capacity divided by design capacity) and the gauge just presents state of charge (current capacity/lastfull)
The better phone charge threshold systems measure usage and keep the phone in the 30-80% soc range as often as possible.
Voltage drops faster on old cells as they age so you need a coulomb counter. Only extremely shit designs guess soc based on voltage alone.
Engineering is compromise though. If you can make a hybrid that loses 5% at 100 but still retains 500wh/l you’re in good shape.
There was someone working on a membrane a while back that’s pretty good at diffusing the lithium transfer in a way that reduces dendrite formation substantially, for instance. That’ll drop your volumetric advantage and likely your max discharge and charge rate a bit but would fix a lot of other problems in the bargain.
I’m not saying that the solution, but there is a palette of tools you can mix and match and that may be one of them.
But does it keep dropping? Is it 60% at 200 cycles
I saw a video on the CATL sodium batteries the other day and the deal is that they’ve found a way to reinforce the material in a way that brings up the slope of the back half of the discharge curve so it’s almost as good as lithium down to about 20% state of charge before falling off the cliff. Lithium is more like 10% but that’s something you can manage with charge circuitry and overprovisioning.
So yeah I’d like to know the answer to your question too.
Going down 10 times faster seems like a really bad trade off for 2 times the capacity. That means your battery will only lst 1/5 or 20% as long.
Not if your application requires 2X the energy. Aircraft, drones, etc. There's always trade-offs in battery design. As an old saying goes: you can have high specific energy, low degradation, or low cost... pick two!
Charge cycle capacity drops are generally not linear. If we start with 2x capacity and drop to 1.6x after 100 cycles, then we might end up with 1.2x after 1000 cycles. Some smartphone manufacturers would love that as you start with extremely superior energy density and then have a built-in obsolescence.
And how much commercial development have NMC and LFP batteries had since they left the laboratory?
LFP batteries are currently being used in newer EVs, most larger power banks, and in newer high-end phones and laptops.
LFP are common in EV’s and ‘solar generator’ style battery packs, but I’ve never seen them in phones or laptops (outside OLPC), reduced capacity makes them not great in these, better to go NMC.
I think this would be perfect for race cars. We might be getting closer to a serious EV endurance series.
Perfect for kamikaze drones probably
You can get a lot of energy density by making the batter non-rechargeable.
See https://en.wikipedia.org/wiki/Metal%E2%80%93air_electrochemi... in general.
A car with this battery can easily have a 1000-mile range (a real one, not EPA). So 100 full cycles would still mean 100k miles!
Yeah, exactly. After your first 100 cycles you only get 800 miles per charge, which is still way beyond what you need for EVs to replace all ICE cars.
> "That's really terrible."
Not really. At 1270 Wh/L, even with 20% degradation, these cells still retain far more energy than a LFP cell (which are more like 350 Wh/L).
The question is, what happens at 200, 500, 1000 cycles? Does the degradation continue linearly or does it slow down? ... or accelerate?
No one knows, the paper just focused on 100 cycles, but it suggests that if its good at 100 it probably is not terrible at further cycles. I guess we'll have to wait for the next paper but the conclusion seems optimistic about future research:
It is important to note that additional improvements in practical cell parameters, such as further optimized electrolyte (E/C ratio), increased stack pressure, optimized separator selection, and higher areal capacity of cathodes, can potentially enhance both the energy density and cycling performance beyond laboratory-scale demonstrations.
Post-mortem analyses confirmed reduced Li accumulation, minimized swelling, and suppressed cathode degradation, validating the robust interfacial stability of the system. By concurrently addressing the reversibility of Li metal and the structural stability of Ni-rich layered cathodes, this synergistic design offers a scalable and manufacturable pathway toward high-energy, long-life anode-free LMBs.
> It's interesting, but 20% loss after 100 cycles is just not great. NMC gets that at near 1000 cycles. LFP gets that at near 5000 cycles.
NMC and LFP had similar issues when these chemistries were at laboratory scale. Give it time and the issues will be solved.
Can the liquid be agitated to avoid dendrite growth?
The schematic images are misleading. In reality, the separation between electrodes is usually on the scale of 1mm at most.
Ultrasonic agitation? Or vibration?
A battery cell is a long thin ribbon that is rolled into a spiral shape. There's no way you can apply any mechanical agitation to all the layers. It's been tried, but nothing came out of it.
That’s fine but it’s only for the first bunch of cycles, after that it’s way worse than standard lion batteries.
Things get better as the technology gets more mature. It's a promising start for sure.